Control theory and artificial intelligence integration in manufacturing and supply chain management: conceptual synthesis and perspectives

Control theory (CT) offers rigorous mathematical methods and models for stability, feedback, and performance analysis. Artificial intelligence (AI) is also rooted in the principles of feedback, stability, and performance control, but views them from a different perspective: model-agnostic, data-driven adaptability and predictive capabilities. This paper reviews CT applications and the mutual enhancements between CT and AI in manufacturing and supply chain operations. Based on literature analysis, we develop a learning-control framework based on a conceptual synthesis of CT and AI and grounded in feedback control, dynamic optimization, and learning-based adaptation. Generalized mathematical formulations illustrate how classical control concepts can integrate with AI-based models, including machine and reinforcement learning, multi-agent systems, and agentic AI. Potential barriers and future research areas are discussed. The paper concludes with implications for designing next-generation industrial and supply chain systems capable of learning, adapting, and stabilizing in uncertain and data-rich environments in the AI era.